Risk factors for violent behavior among incarcerated male psychiatric patients: a multimethod approach.

A multimethod approach that included demographic, criminal offense, drug use, neuropsychological, Rorschach, psychiatric diagnosis, and psychopathy characteristics was used to evaluate 131 incarcerated male psychiatric inpatients. Each criminal offense was ranked from nonviolent to severely violent, and participants were classified as having lifetime histories of either high or low violent behavior. Univariate analyses revealed 12 characteristics which independently discriminated inmates with lifetime histories of high or low violent behavior (married, non-Caucasian race, Axis I psychotic diagnosis, drug other than alcohol or marijuana used most, positive for psychopathy on the Hare Psychopathy Checklist Revised, scores indicating impairment on the Halstead Impairment Index and Category Test, and five Rorschach measures). Logistic regression revealed that eight of those characteristics (married, non-Caucasian race, Axis I psychotic diagnosis, positive for psychopathy on the Hare Psychopathy Checklist Revised, scores indicating impairment on the Halstead Impairment Index, and Rorschach Coping Deficit Index, Personal Responses, and Raw Sum Special Scores) significantly contributed to identifying inmates with lifetime histories of highly violent behavior. Data are provided for sensitivity, specificity, positive and negative predictive power, and overall correct classification rate for neuropsychological and Rorschach measures. Because this sample represented only incarcerated male psychiatric inpatients, and not a general prison population, caution as to limits of generalization are discussed. Implications for use of this information in understanding violent behavior are also discussed.